June 06, 2010
The spinning progress indicator in Photoshop CS5
I’ve heard a number of questions about the new spinning progress cursor (screenshot) that Photoshop CS5 uses. The cursor simply replaces the 1984-vintage MacApp watch cursor (non-standard in Cocoa), for which Apple provides no modern replacement on OS X.
Pierre Igot and others are mistaken in thinking that the CS5 cursor is an alternative to (or, more damningly, an attempt to hide) the “spinning beachball of death.” Photoshop uses the beachball when it’s warranted–i.e., when the app is unresponsive. (The beachball is provided by the OS when the app doesn’t process user interface events fast enough.) It has always used the watch cursor in other cases, where the app is busy performing a task but isn’t locked up, and where the task should be done shortly.
Let’s say we’d (inappropriately) started using the beachball in all cases where we’d previously used the watch cursor. Though it would have no impact (positive or negative) on performance, it would have a big impact on perceived responsiveness, and we’d start hearing “CS5 locks up all the time!” This would be especially profound given persistent misperceptions (arguably deliberately cultivated) of what 64-bit means.
Let’s say that instead of using the watch cursor, we’d pop up progress bars all the time, where none existed in the past. Same result: same performance, unhappier customers.
Let’s say we’d stuck with the 1984-era watch cursor (which I saw pop up the other day in Safari). We’d hear “Look, PS is still Carbon/32bits in places!” You don’t think so? I actually have commenters claim that the all-caps text in iTunes (and previously Finder) is somehow due to Carbon. (And arguably, given the decline in society’s use of wristwatches, I can imagine catching flak about relying on a floppy disk-like anachronism.)
We gave this issue a lot of consideration, and ultimately there wasn’t a perfect solution, so we chose a route that modernized the app UI without misusing standard Apple UI or misrepresenting app responsiveness. Maybe it’s something we can improve in the future, and we’re certainly open to feedback.
Of course, the real issue here has very little to do with one cursor vs. another. It has to do with a question of whether Adobe builds “real” Mac software. We do, and we’re making more progress all the time. There’s a lot of detail to unpack here, and being on the road, I can’t get into all of it now. I’ll try to do so soon, and in the meantime you might want to see my thoughts about platform consistency.